From: Tom Lane <tgl@sss.pgh.pa.us>
To: David Fogarty <dfogarty@icsusa.com>
cc: pgsql-novice@lists.postgresql.org
Subject: Re: Performance after restart/reboot pre-Analyze
In-reply-to: <6F39634B-20B1-4683-9320-C4D7B68D53CF@icsusa.com>
References: <6F39634B-20B1-4683-9320-C4D7B68D53CF@icsusa.com>
Comments: In-reply-to David Fogarty <dfogarty@icsusa.com>
	message dated "Thu, 28 Mar 2024 10:43:37 -0500"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3584162.1711641179.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Thu, 28 Mar 2024 11:52:59 -0400
Message-ID: <3584163.1711641179@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3584163.1711641179%40sss.pgh.pa.us>
Precedence: bulk

David Fogarty <dfogarty@icsusa.com> writes:
> It seems like whenever I restart my database server performance is quite=
 bad until an Analyze has been performed. I expect this is due to the quer=
y planner needing new stats, but that leads me to three questions:

A simple restart shouldn't invalidate statistics.  I'd lay some small
change on your actual problem being not that, but the need to
repopulate the server's buffers from disk files.  If so, pg_prewarm
might help you.

			regards, tom lane